Whether it is OEM professional OEM, ODM original factory commissioned design, or even outsourcing, there is a common point in the modern industry. What is the common point? First of all, the factory produces products. The most important thing is the yield rate of the product, that is, the percentage that falls within the specification. However, if the quality of the product is to be evaluated, the yield rate is not the only indicator. The most common evaluation is the quality of the process capability, such as CPK>1.33, etc., but I know why big companies such as Google, Intel, Apple can make the product quality stable, which involves the tools and indicators they use for evaluation. The following will introduce a simple SMT case and use JMP for analysis .
In the SMT (Surface-mount technology) process production line, one of the site glue processes requires evaluating the quality of equipment from two different manufacturers and selecting the appropriate machine. Before the analysis, let everyone briefly understand the dispensing process.
In the dispensing process, the thermal expansion coefficients of the wafer and the substrate are different, and the temperature changes cause thermal stress to concentrate on the bumps between the two, causing cracks to form an open circuit (Figure 1). When the substrate is deformed by external forces or the environment is oxidized, the bumps will also change. Brittle deterioration, so it is necessary to dispense to increase the bonding strength between the chip and the substrate, isolate the air and redistribute the stress (Figure 2). Of course, there are many factors in the dispensing process, this time it is only simple Evaluate the weight consistency of the machine after dispensing and the degree of deviation in the x and y directions.
Picture 1 on the left; Picture 2 on the right
The technical specifications usually provided by the machine are shown in Figure 3 below (not the standard for this example)
image 3
01: Descriptive statistical analysis and comparison
When analyzing, we start with the dispensing weight. The data collection includes company (M, W) and number of days data (Figure 4 below). It is recommended to start with Descriptive statistics To start, understand its distribution first, and enter the specifications for comparison.
Figure 4
From (Figure 5 below), we can see that the two companies are not very symmetrical. It can be easily compared using the Distribution and Graph Builder platforms in JMP.
Figure 5
Then match the normal state verification (see Figure 6 below) to determine it as Abnormal allocation , So we can then compare the process capabilities,
Image 6
The process capability of M company is higher than that of W company, but all have a level greater than 1.33 (Figure 7)
Figure 7
02: Hypothesis test
After comparing the narrative statistics, you can then use the hypothesis test to find out whether the two companies are significantly different. From the test results, it is known that the variation between the two is not equal (because both of them are abnormally distributed and use the Levene method to judge), Then the Wetch Test understands that there is a significant difference between the two mean values, so in the weight part, it can be clearly seen that M company has an advantage, as shown in Figure 8.
Figure 8
03: ANOVA analysis
The next step is to look at the number of days to understand whether different days will cause significant differences in the weight of the two companies. According to the collected data, perform ANOVA analysis on different days, as shown in Figures 9 and 10 below. There is no significant difference in some parts, and there is a significant difference in the mean part. Assuming that the deviation does not affect the quality, the company will generally set another Practical difference as the level, depending on whether it will actually affect the quality. However, company W obviously has a significant difference in the two-day level, even if the mean is not significantly different, in general, it is more difficult to adjust the variation than to adjust the deviation.
Figure 9 - ANOVA analysis of M company
Figure 10 - ANOVA analysis of W company
04: Control Chart
In addition to statistical verification, it can also be used with the control chart function (Figure 11), with Local Data Filter to filter and compare data. It can be seen from the graph that the two equipment vendors have a narrower control boundary on the second day, but W company is very clear There are points close to the specifications. If it is really due to price considerations, the reason must be investigated.
Picture 11
Then confirm the degree of deviation of xy, and confirm according to the above process. It is confirmed from the process capability (Figure 12) that Company M is better, and the two companies perform ANOVA analysis (Figure 13) whether it is a variation or a mean value. There are significant differences between equipment vendors.
Picture 12
Figure 13
Then subdivide it into two days for comparison (Figure 14). Only the differences are listed here. Whether company W is in the x or y direction, the number of days will cause differences, while company M has no effect.
Figure 14
Finally, with the control chart confirmation (Figure 15), confirm the level of W company, and M company will not put the above chart if there is no significant difference, so whether it is the weight of the dots or the offset confirmation of xy, it is better for M company.
Figure 15
in conclusion
Finally, from the perspective of the client, there are many quality parameters to be confirmed for just one process. In addition, each parameter also has different methods to analyze. Although the above functions are scattered on different platforms, they can be customized with JSL. Windows or Dashboard integrates all analysis results and standardizes them, echoing the common point mentioned at the beginning is "quality and timeliness of analysis". There are many applications in different industries in real life. Welcome to register for the JMP online seminar. , You will hear more practical analysis scenarios to help you solve analytical problems in daily work.
Latest online seminars // 6 examples of data science in high-tech applications
In complex high-tech production lines, to effectively enhance product competitiveness and optimize manufacturing processes, the use of data to discover, solve and prevent problems has become an important key that cannot be ignored. This seminar will start from the perspective of production, through the sharing of practical application cases, understand how JMP can effectively reduce unnecessary working hours and improve work efficiency; then learn and avoid losses, try to use the most practical examples and the most efficient Methods to solve and discover problems, further optimize production conditions and make early warning and timely decision-making. This is a free seminar, Click here to register now
